Abstract: In sensorless control of brushless DC motors using back electromotive force for commutation timing, traditional sensorless control systems often experience a lag in zero-crossing detection, leading to a decrease in the accuracy of commutation in the control system. When the ratio of the sampling frequency and the frequency of back potential force is low, the commutation failure is easy to occur. In response to this issue, a novel approach for determining the commutation timing is introduced, and a fuzzy control algorithm is employed to enhance the precision of commutation. Finally, STM8003F is used as the main control chip to build an experimental platform to compare and analyze the control performance based on the improved phase change algorithm and the phase change algorithm based on the inverse potential, and the experimental results show that the improved phase change algorithm has the advantages of shorter phase change delay, more stable phase change, and smaller amplitude of current fluctuation.